HAL Id: hal-02598783
https://hal.inrae.fr/hal-02598783
Submitted on 16 May 2020
HAL is a multi-disciplinary open access
archive for the deposit and dissemination of sci-entific research documents, whether they are pub-lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.
L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.
Influence of the forest composition on bryophytic
diversity in mixed and pure pine and oak stands
D. Fourcin
To cite this version:
D. Fourcin. Influence of the forest composition on bryophytic diversity in mixed and pure pine and oak stands. Environmental Sciences. 2013. �hal-02598783�
UNIVERSITE D’ORLEANS
COLLEGIUM SCIENCES ET TECHNIQUES ITP SCIENCES BIOLOGIQUE ET CHIMIE DU VIVANT
MASTER SCIENCES BIOLOGIES (2ème année) SPECIALITE : BOPE
Année 2012 – 2013
Deki Fourcin
Irstea - UR Écosystèmes forestiers - Domaine des Barres - Nogent-sur-Vernisson
1
Influence of the forest composition on bryophytic diversity in
mixed and pure pine and oak stands
Deki Fourcin, under the direction of Marion Gosselin
Irstea - UR Écosystèmes forestiers - Domaine des Barres - Nogent-sur-Vernisson
Abstract
We studied the specific richness (SR) and abundance of epiphytic and ground floor bryophytes in mixed and pure stands of sessile oak (Quercus petraea) and Scot’s pine (Pinus
sylvestris) in the Forest of Orleans (France). Stand and tree level assemblages were analysed
using general linear modelling in 21 stands of mixed or pure composition.
Results showed that SR and abundance was mainly determined by the host tree species, and we observed higher SR on oak trees than pine, and on pine trees in mixed populations compared to pure. Abundance was found to be higher in pure populations for both oak and pine. An analysis of the composition is necessary in order to interpret these preliminary results.
In terms of forest management our results show that pure and mixed composition has no effect on the SR at plot level. Mixed populations are beneficial for the biodiversity at tree-level. Addition of covariates in the models gave results that allowed us to form new hypotheses. Most influential covariates seemed to be the chemical properties of throughfall and stemflow, followed by measures implicating stand density, which had an effect with the plot type on the diversity.
Key words: epiphytic bryophytes, forest structure, ground floor bryophytes, Pinus sylvestris,
Quercus petraea
Résumé
Nous avons étudié la richesse spécifique (RS) et l’abondance des bryophytes (epiphytiques et terricoles) dans des placettes pures et mélangées de chêne (Quercus petraea), et pin (Pinus
sylvestris), dans la Forêt Domaniale d’Orléans (France). Les assemblages étaient analysés
aux échelles placette et arbre en utilisant des modèles linéaires généralisés sur 21 placettes de composition pure ou mélangée.
aijijiiiiiiLes résultats montrent que la RS et l’abondance sont déterminées en majorité par l’essence hôte, et nous avons observé une RS plus élevée sur chêne que sur pin. L’abondance des terricoles était plus élevée dans les populations pures que mélangées. Toutefois, il serait nécessaire d'analyser les données en composition pour pouvoir interpréter ces premiers résultats.
En termes de gestion forestière, ces résultats montrent qu'à l'échelle du peuplement, son caractère n'entraîne pas de différence significative de richesse ou abondance des bryophytes associées à une essence donnée. Les peuplements mélangés sont un plus pour la biodiversité des bryophytes à l'échelle de l'arbre-support. L'ajout de covariables nous permet de formuler de nouvelles hypothèses. Les facteurs les plus influents semblent être les propriétés chimiques des eaux d’écoulement et des pluviolessivats, ainsi que la densité des peuplements, qui interagit avec la type de peuplement sur la diversité.
Key words: bryophytes épiphytique, bryophytes terricole, Pinus sylvestris, Quercus petraea, structure forestière
2 1. Introduction
In 2007, the Grenelle de l’Environnement defined the goal of increasing the use of renewable energy by 23% before the year 2020, thereby increasing the exploitation of wood. It is important to base forest management processes on not only the yield, but also the conservation of biodiversity. Vallet (2011) found that mixed species stands could well be the answer, as forest productivity and biodiversity are possibly favoured. Numerous studies looking at the biodiversity are underway, and this work is part of the Iscar project (Indicators of Structure, Composition and Analysis of ScenaRio), which analyses the effects of varied stand composition in lowland temperate oak and pine forests on different groups of biodiversity indicators.
We will be looking into the effect of mixed versus pure stands on the bryophyte diversity (epiphytic and ground floor), and field work will be conducted on mixed and pure populations of Scot’s pine (Pinus sylvestris) and sessile oak (Quercus petraea) in the Forest of Orléans. Bryophytes are important for overall biodiversity, as they are primary producers, and serve as essential food resources (Virtanen et al. 2000)
As primary producers, bryophytes contribute to overall biodiversity, and serve as essential food resources, for example, for mammals (Virtanen et al. 2000) and habitats for molluscs and insects. They are also important indicators of invertebrate diversity (Cleavitt et al. 2009a). Their diversity is affected by the macroclimate, and to a greater extent, the micro-climate (Raabe et al. 2010a). Regionally, forest continuity and management practices are understood to be the main factors affecting bryophytic composition (Aude and Poulsen 2000).
We aim to find the impact of stand composition (pure pine, pure oak, and mixed pine-oak) on the epiphytic (bark inhabiting) and ground floor bryophyte diversity present on each host tree species. Our study also concentrates on the influence of individual tree species on bryophytic diversity, including their associated physio-chemical influences, such as pH of the substrate (soil and bark), and water supply (throughfall and stemflow).
There are limited publications with a direct link between oak-pine mixed stands, and varying bryophyte diversity on individual tree species. However, diversity in relation to tree composition and forest management practices has been widely discussed (Farmer et al. 1991; Márialigeti et al. 2009; Raabe et al. 2010; Weibull et al. 2005; Wallrup et al. 2006; Humphrey
3
et al. 2002; Cleavitt et al. 2009), and sufficient data exists in order to form various hypotheses. Most relevant perhaps, is Cleavitt’s study (Cleavitt et al. 2009) which found that the epiphytic bryophyte diversity is higher in mixed maple and spruce stands at both tree and plot level. This result was confirmed by Kiraly et al. (2010) in Hungarian oak and pine forests, and Felton et al. (2010) for spruce and birch. Humphrey et al. (2002) found that epiphytic diversity was higher on oak than pine. Concerning ground floor species, Márialigeti (Márialigeti et al. 2009) found that structural homogeneity is harmful in terms of abundance and species richness, and that the relative density of oak in a pine forest had a positive effect on the diversity of ground floor bryophytes. Koptsik et al. (2001) confirmed that the abiotic conditions produced by broad-leafed trees affect the forest floor bryophytes, as the vascular ground vegetation is often related to the tree layer composition. Mixed stands potentially host a larger range of conditions, and consequently a greater amount of bryophyte species.
Although the tree species is important, on a local scale bryophytic diversity is explained by substrate structure and chemistry (Cleavitt et al. 2009), light, wind and precipitation exposure, and support tree species (Aude and Poulsen 2000). Trunk diameter and bark traits (water retention, scaling and texture) also play a role (Cleavitt et al. 2009). Some articles (Gregory et al. 2002; Peck 1997; Slack 1976; Schmitt et al. 1990; (Cleavitt et al. 2009), go as far as to say that for epiphytic diversity, climatic and edaphic factors are far more important than the tree species. In order to explain the variations in diversity levels associated with different trees, we drew up a list of intervening variables that could interact with the host tree species or plot type.
Bryophyte species have an optimum pH, which suggests that they are distributed along gradients of substrate chemistry and moisture regimes, and are not actually specific to given tree species ((Cleavitt et al. 2009), for epiphytes). Although germination of spores can take place in a wide pH range, the protonema (first development stage) range is more restricted (Glime 2007), especially under acidic conditions (Fritz et al. 2009). The pH of the environment affects the bryophyte’s enzymatic activities, the germination of some species, and influences certain major functions within the plant, such as the photosynthesis (Glime 2007).
Bryophytes assimilate a small amount of water from their substrate (Raabe et al. 2010), and to a greater extent with their entire above ground surface. We therefore need to observe the
4
substrate pH (bark and soil), as well as the throughfall (rainwater that passes through the tree canopy and therefore supplying the ground floor bryophytes) and stemflow (rainwater which runs down the trunk epiphytic bryophytes). In general terms, the number of species able to survive under acidic conditions is limited (Barkman 1958). The bryophyte must be able to tolerate the whole range of pH levels in order for it to survive, rendering it an important factor affecting bryophyte diversity. Oak and pine trees differ in soil acidity preference, oak tolerating pH 3.8 – 6 (Farmer et al. 1991), compared to pine’s pH 3.4 – 4 (Barkman 1958; Kuusinen et al.1996), which could explain the lower diversity associated with pine trees.
The leafed canopy of deciduous species is known to neutralise the acidity of incident precipitation (Fiedler et al. 1981; Cronan 1983; Puckett 1991), which will result in less acidic stemflow, and throughfall. However, stemflow on pine trees collects acid compounds from dry and wet atmospheric deposition, potentially acidifying the bark (as seen by Fritz et al. (2009) on beech trees). The smooth nature of pine bark, similar to that observed for beech by André et al. (2008) is also more conductive to stemflow, increasing the negative effect for the bryophytes on pine. Augusto et al. (2001) also found that Scot’s pine trees intercept almost double the precipitation volume compared to sessile oak. Barbier et al. (2009) confirmed these results when comparing deciduous and evergreen trees during both the growing period and over the whole year.
Previous articles have considered the mean stemflow pH for different stand types, and their results suggest a decreasing gradient of acidic effects of stemflow function of the distance from the tree (Beniamino et al. 1989; Vellak et al. 2003; Cleavitt et al. 2009; Chang et al. 2000). As well as pH measurements, we decided to effectuate conductivity readings on all samples. Conductivity of water is known to be related to bryophyte diversity in aquatic habitats (Ceschin et al. 2012), and is used as an indicator of the nutrient levels in the water supply. As yet, there are no publications linking the conductivity of the water supply for non-aquatic bryophytes, although some studies have been done in the case of bark and water supply conductivity for epiphytic lichens (Hauk et al. 2005).
Other factors which could intervene in the bryophytic diversity involve the density and trunk diameter of trees in the stand. A large basal area (BA), and consequently elevated relationship between the average Diameter at Breast Height (DBH) and tree density is expected to increase the epiphyte diversity(Aude and Poulsen 2000; McGee and Kimmerer 2002; Bardat et al.
5
2007; (Márialigeti et al. 2009), (Kiraly et al. 2010), and have a negative effect the ground floor diversity (Márialigeti et al. 2009). The RDI is potentially a more reliable measure for stand density as it takes into account the maximum viable stand density in relation to the tree species. Structural diversity (a large standard deviation of DBH) is expected to increase the ground floor bryophyte cover (Márialigeti et al. (2009), Kiraly et al. (2010)).
The sampling plan was constructed to accommodate the results of the literature review, and in order to be able to test the main hypotheses (Table I) which were based on the effect of the host tree species (pine and oak) and the stand type (mixed or pure).
2. Materials and methods
2.1. Study area and sampling plan
The study was carried out in the Forest of Orléans in central France. The forest is temperate, and stretches over 60km to the North of the Loire, between Orléans and Gien, covering around 35,000 hectares, and an elevation of between 107 and 174 m above sea level. The majority of the tree flora is Scot’s pine and sessile oak, interspersed with hornbeam, birch, and to a smaller extent beech. Both oak and pine are temperate species, Scot’s pine being native to almost the whole of the northern hemisphere, and sessile oak confined to Europe, Table 1 Main hypotheses to be tested
Bryophyte
Character Tree Level Plot level
H1 Specific richness will be significantly lower on pine compared to oak.
H7 Specific richness will vary in function of the forest type (mixed or pure).
Epiphytes
H2 Specific richness on oaks will vary in function of the forest type (mixed or pure) in which the support-tree is positioned.
H3 Specific richness on pines will vary in function of the forest type (mixed or pure) in which the support-tree is positioned (higher species richness and abundance in mixed populations).
H4 The species richness and abundance within 1m of a pine tree will be significantly lower than that of oak.
H8 Specific richness and abundance will vary in function of the forest type (mixed or pure).
Ground floor
H5 The species richness and abundance 1m of an oak will vary depending on the stand structure of the forest. Species richness and abundance will decrease in mixed stands.
H6 The species richness and abundance within 1m of a pine will be higher within a mixed population, and composition will be different.
6
excluding the extreme north and south. Field work was based in the far Eastern section in the Lorris area.
Plot choice was based on the original hypotheses (table 1), and the aim was to use a balanced selection of different plot types in order to express the bryophyte diversity. All plots were situated in the same geographical area, and consisted of seven plots for each level of the main explanatory variables: pure pine, pure oak, and mixed oak – pine populations. Each plot was circular, and had a radius of 15m. Plots were split into three equal segments, and the largest tree(s) (largest oak and pine for the mixed populations) in each sector was chosen for the bryophyte analyses. We therefore had three sampling trees for each pure plot, and six in each of the mixed.
2.2 Data Collection for Bryophytes (dependant variables)
Due to an expected decreasing gradient of acidic effects of stemflow function of the distance from the tree (Beniamino et al. 1989; Vellak et al. 2003; Cleavitt et al. 2009; Chang et al. 2000), we decided to do ground floor bryophyte inventories at two different distances in order to monitor the potential effect. Field work included data collection for all three aspects of bryophyte diversity (species richness, abundance and composition), although this article will only be treating the species richness of epiphytic bryophytes, and species richness and abundance for ground floor bryophytes. On top of the observations of bryophytes we
7
conducted measures for the secondary variables, although it must be noted that the sampling plan is not necessarily balanced for the range of values for these results.
a) Ground floor bryophytes
Bryophyte diversity was recorded over an area of 50 x 50 cm² in three directions (azimuts 167, 300 and 333 gr) at two distances from the tree (25 and 75 cm). Data for each distance was combined, giving us a composite result for bryophyte species and abundance at 25 and 75 cm from the trunk. An inventory as complete as possible was effectuated for each quadrat, followed by an estimation of abundance (in cm²) using a graded system.
b) Epiphytic Bryophytes
An inventory of all bark dwelling species took place on the entirety of each sample tree trunk up to a height of 2m. For the measures of abundance we evaluated both the total epipiphytic cover and the proportion of each bryophyte species in relation to the bark surface. Abundance was noted as follows:
- Class 0: Absence of bryophytes - Class A: < 1 % cover - Class B: 1 % ≤ Cover < 5 % - Class C: 5 % ≤ Cover < 10 % - Class D: 10 % ≤ Cover < 25 % - Class E: 25 % ≤ Cover < 50 % - Class F: Cover < 25 %
2.3 Data Collection for Covariates
a) Substrate i) Soil
A 2.5cm deep soil coring ring was used to collect samples of both the fragmented organic (OF) horizon, and the organo-mineral (A) horizon from each sampling tree. Samples were taken at 25cm at 75cm from the trunk in three directions (same placement as bryophyte observations), then combined in one envelope to create two composite samples for each tree. In the end we had four envelopes, each containing three soil samples. Envelopes were labelled with the soil type, as well as the date, site, plot and tree numbers and distance from trunk. norm ISO 10390 (AFNOR, 1996)
Pre-analysis in the laboratory involved the choice of the protocol for pH and conductivity measures on the soil. We chose to follow the norm ISO 10390 (AFNOR, 1996) method as this seemed to be the most universal. Samples were first air dried for two weeks, and then sieved using a 2mm square-holed sieve. 5ml of soil was added to a beaker along with 25 ml
8
demineralised water, and then agitated for one hour using a magnetic stirrer. The solution was then left for 30 minutes so that the sediment would settle and thereby allow us to use a filter syringe (filter size 0.45 µm) to transfer the solution into a clean beaker. This was done to prevent the organic layer from the soil from damaging the probe. We then measured the pH of the solution using a pH metre (Eutech pH 6+ Meter) and conductivity of the solution using a conductivity metre (Eutech Cond 6+).
ii) Bark
A coring ring was used to take a 4cm² bark sample from the West side of each tree at 1.10 m. A chisel was used to ensure that only the outer surface of the bark was removed, and bark samples with a minimum amount of lichen were favoured. Samples were stored in plastic sachets and left to air dry.
Pre-analysis was necessary to determine the protocol to be used for the the measurement of bark pH and conductivity. Most methods reviewed (Bates et al. 1981; Mitchell et al. 2005; Cleavitt et al. 2011) involved grinding samples, although Asta et al. (1998) state that this does not significantly change the pH result, and Farmer et al. (1990) state that this is unnecessary as we stay closer to the natural environment by leaving the sample more or less intact. The same argument goes for the presence of lichen, as it forms part of the moss habitat. Many studies (Cleavitt et al. 2011; Fritz et al. 2009; Hauck et al., 2005; Legrand et al. 1995) remove any lichen present as they believed that this would affect the pH, however, we verified that pH for samples from the same tree did not differ significantly if lichen was removed. Further testing was effectuated in order to decide on the best soaking time for the samples.
The outer layer of the bark with lichen intact was broken into small pieces using a chisel, and then 0.4g of bark was added to 4ml of demineralized water. The sample was left to soak for one hour before reading the pH using a pH meter. Before taking the conductivity reading, a further 15ml of demineralised water were added in order to make the solution depth sufficient for the size of the probe.
c) Water Supply
Two series of water supply samples were collected: One at the beginning of October, when the canopy was leafed, and the second in February in the deciduous non-leafed period.
9
Throughfall is the rainwater that falls to the ground through the canopy, and stemflow is rainwater that runs down the trunk of the tree.
An ‘open field’ cup placed in open area for the duration of the water supply collection in order to have an idea of rainfall volume and consequently the proportion that is transformed into potential water supply for the bryophytes.
i) Throughfall
Before a rainy episode a plastic pot was anchored in the soil at both 25 and 75cm from the tree on the West side. Each cup was covered by a grill which was attached using wooden pegs, to prevent any substrate from falling into the pot. After rainfall the cups were collected, and labelled with the date, site, plot and tree numbers, and distance from the trunk. Samples were stored in the refrigerator, preceding pH and conductivity measures for each sample in the laboratory.
ii) Stemflow
Before a rainy episode we attached a plastic cup to each tree that had undergone a bryophyte inventory in order to collect the stemflow. A stainless steel blade was inserted horizontally on to the bark on the West side of the tree at a height of 1m. The blade was manipulated in a way as to form a slight groove in the middle, and direct the water into the cup which was attached to the tree using a length of wire placed just underneath. A Cup was attached just underneath each blade by twisting a loop in the middle of a length of wire to hold the cup, then fastening the ends of the wire around the tree. Cup lids were attached loosely and then covered with tin foil to prevent overhead water from being collected. After rainfall the cups were collected, labelled, and then stored in the refrigerator. pH and conductivity measures were effectuated for each sample in the laboratory.
d) Density
The telemeter Vortex II was used to measure the distance of all trees from the centre point of each plot. A tape measure for circumference of trees, and particularities concerning the trees were noted. Raw data for trunk diameter and stand density was collected in order to calculate basal area (BA) and Relative Density Index (RDI) of each plot.
10 a) Pre-analysis
Preliminary analysis consisted of checking results series for outliers, followed by the generation of histograms for all data sets to verify the normality. We generated correlation matrices to compare series of measures for covariate data. Coplots were then created to test possible interactions between the explanatory variables and the covariates.
b) Testing the main hypotheses
We started by generating simple models for each hypothesis, for which the sampling plan had been constructed. This included data for the response (bryophyte diversity specific richness and abundance), and the explanatory variable (plot or species).
i) Specific Richness
Data collected followed a discrete probability distribution, and we therefore used either the Poisson or quasipoisson distribution for our modelling. Depending on the hypothesis being analysed, we used either the plot type or the tree species as the explanatory variable.
When testing tree species effect (H1 and H2), we were looking at pine and oak trees within the same stand. We worked on the basis that these models could be generated using a Generalised Linear Model (GLM) , as there was no effect ‘plot’ to consider. The Quasipoisson law was chosen in order to account for the under-dispersion of our count-data results.
For the plot type effect (H2, H3, H5, H6, H7 and H8) we generated models using Generalised Linear Mixed Models (GLMM), in order to take the effect plot created by the geographical variation into account. We were obliged to use the Poisson law even though our data was under-dispersed, as Quasipoisson is not possible for GLMM. Models for hypotheses which showed no random plot effect (H2, H3, H4, H5 ground floor 25cm, and H6) were regenerated using a GLM and the Quasipoisson law, which was better suited to our under-dispersed data. This left us with GLMM models for just H5 ground floor bryophytes at 75cm, H7, and H8.
ii)Abundance
Models of species abundance were generated using the binomial law, and the logit link function. This was due to the quantitative nature of our abundance data which were limited between fixed values. We incorporated random plot effects as necessary. Epiphytic bryophyte
11
abundance was not taken into account due to the nature of the data and lack of time: ordered classes such as those obtained are difficult to analyse without using Bayesian statistical modeling.
c) Addition of covariates to models
In order to see if improvements could be made on initial models, we chose potentially interesting covariates and incorporated them into those which gave significant results. The criteria used for comparing models was the Akaike Information Criterion (AIC), or the the qAIC for models using the Quasipoisson law. Comparisons of models were done two by two when treating data sets with non-available data as it was necessary to use data sets of the same length. These models were used to generate further hypotheses and suggest potential paths for further work.
3. Results
3.1 Original Hypotheses
a) Tree species effect
i) Epiphytic bryophytes
SR of epiphytes was significantly lower on pine trees compared to oak trees (‘H1’, Table 2), with a mean difference of 1.2 species. A more significant result was observed for the SR between pine trees in pure and mixed plots (‘H3’, Table 2), although the difference in SR was also 1.2. No difference in SR was observed between pure and mixed plot oak trees (H2).
ii) Ground floor bryophytes
In terms of significant differences in diversity between oak trees and pine trees in mixed stands (‘H4’, Table 2), we found a significant difference in SR of ground floor bryophytes at 25cm from the trunk. The model result was less significant however, than that for epiphyte SR. No differences were found for SR of ground floor bryophytes at 75cm, or abundance at any level.
SR did not vary significantly between pure and mixed oak populations (H5), for either bryophytes at 25 or 75cm. For differences in abundance (‘H5’, Table 2) there was a large variation in ground floor bryophytes at 25cm between plot types. The bryophyte cover in mixed populations was more than double that found in pure oak plots. Differences in abundance at 75cm from the trunk were not significant.
12
(a ii) (a i)
(b ii) (b i)
Fig. 3 Total Specific richness (i) and abundance (ii) at tree level (a) and plot level (b). SR stands for specific richness; AB stands for abundance; Ground 25 stands for ground floor bryophytes at 25 cm from the trunk; Ground 75 stands for ground floor bryophytes at 75 cm from the trunk.
13
Significant differences for diversity between pine trees in pure and mixed plots (H6) were found at two levels: SR was higher for ground floor bryophytes at 25 trees in mixed stands, and abundance for bryophytes at 75 cm was higher in pure stands.
Significant models show that abundance and SR are not affected in the same way by the explanatory factors: SR was affected negatively by pine trees in pure stands compared to mixed, and by pine trees compared to oak trees, whereas the opposite effect can be observed on the abundance in both cases.
*
b) Plot level
Boxplots illustrating bryophyte diversity at tree level are illustrated in Fig.3. (bi) for all types of bryophyte populations. Models did not show significant differences between pure and mixed plots for oak and pine trees. This was true for RS of epiphytes (H7), and RS and abundance of ground floor bryophytes (H8) (See Fig.3 for illustration of results).
3.2 Addition of covariates to models
14
The original model for H1, showing a lower SR on pine compared to oak trees was not improved by the addition of covariates (‘H1’, Table 3), with the addition of stemflow pH giving a higher AIC value. Stemflow pH measured on pine trees was more acidic than that on oak trees, which had a negative effect on the SR of epiphytes. However, within the pine and oak tree populations, more acidic stemflow pH led to a higher SR.
The original model for H3 was improved by addiding the effect of the RDI, and the interaction between the RDI and the plot type. The higher RDI value found in mixed pine populations had a positive effect on the SR.
b) Ground floor bryophytes
No significant difference in SR was found for Ground floor bryophytes at 25 cm from the trees, when comparing oak and pine trees (H4). For ground floor bryophytes at 75 cm from the tree, models showing the higher SR on oak trees compared to pine trees (H4) were improved by the addition throughfall pH and conductivity covariates (‘H4’, Table 3). We found three models with lower AICs than the original, the best using just the throughfall conductivity to explain the variation in SR. Influence of the conductivity was more significant than the explanatory variable, and an increase in conductivity had a negative effect on the SR. The second best model in terms of the AIC, also showing a more significant result than the species effect, was the pH of the throughfall. In contrary to the conductivity, an increase in pH values for all trees had a positive effect on the SR. The third covariate model showed a significant effect of both throughfall pH and the tree species. A higher pH at species level had a positive effect on the RS. Pine trees had a lower mean throughfall pH, which accounts for the lower SR. For H4, the tree species did not have a significant effect on the abundance, either at 25 or 75 cm.
When modelling the different abundance values for pure and mixed oak populations (H5), the addition of covariates did not improve the model (‘H5’, Table 3). However, basal area of the stand was the covariate best describing the higher abundance on oak trees in pure plots, and an increase in BA values had a positive effect on the abundance.
The original model showing higher SR of ground floor bryophytes between pine trees in pure plots (‘H6’, Table 2) could not be improved by the addition of covariates, and was therefore best described by the explanatory factor. The best covariate for explaining the higher
15
abundance at 25 cm on pine trees in pure plots was best described by the BA, although the AIC for this model was higher, and the result less significant than the original. A higher BA, as seen in the pure populations, had a positive effect on the abundance.
4. Discussion
4.1 Original Hypotheses
a) Tree level
i) Epiphytic bryophytes
As expected, epiphytic specific richness was higher on oak trees compared to pine trees (H1, see table 1). In terms of the tree species, this could be explained by the physiological differences. Oak bark is rough and mesotrophic (Márialigeti et al.2009), thereby creating
16
sheltered, moist microhabitats which bryophyte propagules can easily adhere to. The comparatively smooth, flaky and less absorbant pine bark (Barkman 1958; Ahti et al. 1977) is therefore less appropriate for a permanent bryophyte cover, as propagules are washed off the bark by the rainfall (Barkman, 1958).
The increase in specific richness on pines in mixed populations compared to pure confirmed the results of previous studies (H3). This suggests that the different microclimate induced by a change in stand type does have an effect on bryophytes associated with pine trees. This result will be discussed further below, in terms of covariate effects.
We expected the specific richness to vary for oak trees in mixed and pure populations (H2). However, no significant difference was found between the two. This suggests that the host tree is the most important factor in the epiphytic bryophyte diversity on oak trees. This corresponds to the results found by Kiraly et al. (2010), who suggests that the reason for this is the ‘mesotrophic, wrinkle-rich bark structure of oaks offers many shady, windproof and moist habitats, where bryophyte propagules can easily adhere to.’
ii) Ground floor bryophytes
SR for pine trees in mixed plots was higher than that in pure plots at 25 cm. This is similar to the findings of Gustafsson and Eriksson (1995) which suggest that the presence of deciduous species (Populus tremula for this study) has a positive effect on the SR. The reason for this could be that some bryophyte species that are normally oak specific are able to associate with pine trees in mixed stands. Analysis of the species composition would be necessary to confirm this idea.
The fact that SR was higher in oak populations than pine at 75 cm suggests that the microclimate produced by the oak trees is more favourable than that produced by pine trees.
We observed a higher abundance in pure populations of both oak and pine compared to that in mixed populations. For bryophytes at 75 cm in pure and mixed pine plots, this could perhaps be due to a negative effect of litter cover introduced by the presence of oak trees. Márialigeti et al. (2009) found that ‘litter covers the suitable substrates, thus inhibiting the development of a bryophyte layer’. The increase in litter cover is likely to have a detrimental effect on the
17
abundance but not SR. The fact that we are considering bryophytes at 75 cm from the host pine could further increase the perturbation caused by oak litter in mixed stands.
b) Plot-level
Absence of significance when comparing pine and oak plots, and pure and mixed stands is surprising. For oak trees this confirms the observations found at plot level: for bryophytes associated with oak trees it is the host tree and not the plot type which is important.
Perhaps data for composition would have better allowed us to find differences at plot level. This would allow us to identify specialist species and thereby construct more conclusive results at this level.
4.1 Models incorporating covariates
a) Epiphytic bryophytes
In terms of the covariates, the pH of the stemflow was the best factor in explaining the specific richness between oak and pine trees (H1). This follows the discussion of Cleavitt et al. (2009), which noted that the pH of the stemflow is an important bark characteristic in terms of epiphytic composition. Stemflow collects acid compounds from dry and wet atmospheric deposition, and may thus acidify the bark (as seen by Fritz et al. (2009) on beech trees). On deciduous species however, the leaves convey basic cations to the stemflow, thereby neutralising the acidity (Pryor et al. 2005), and being less likely to restrict the growth of epiphytic bryophytes. Thin smooth bark (observed for beech by André et al (2009) is also more conductive to stemflow, increasing the negative effect for the bryophyte diversity on pine. Cleavitt et al (2009) found that sites with lower volumes of stemflow have a higher epiphyte biomass.
To explain the higher SR on pine trees in mixed populations, compared to pure (H3), the best covariate for the model is the RDI ‘H3’, Table 3). The positive effect of the RDI is similar to the results of Márialigeti et al. (2009) and Kiraly et al. (2010), although their results were for the BA. They found that an increase in tree diameter has a positive effect on the epiphyte cover, as it increases the number of microhabitats (McGee and Kimmerer 2002).
18
To explain the higher SR of ground floor bryophytes at 75 cm we can revert to some of the ideas discussed above for the epiphytic bryophytes. All improvements to the original model involved incorporating the chemical properties of the throughfall. The third model shows that the lower pH of throughfall for pine trees has a negative effect on the SR. This is probably due to the fact that the oak canopy raises the throughfall pH (Pryor et al. 2005)
In terms of explaining the higher abundances on oak trees in pure stands than mixed oak (H5) at 25 cm, BA was the best covariate, although the effect was less significant than the effect ‘plot type’. A higher BA had a positive effect on the abundance. This is contrary to the conclusion made by Márialigeti et al (2009), who found that ‘large trees had basically negative effects on forest floor bryophytes’. However, this was in reference to large beech trees. Again, we are in need of information for species composition. It is possible that in the case of our study, specialist species associated with oak were highly sensitive to their microclimate. As seen for other models, the physiochemical properties of the throughfall are likely to have an important effect on the bryophyte diversity. A higher density of oak trees would increase the neutralising effect of the oak canopy and perhaps have a positive effect on the abundance. To test this theory, we could construct models containing more than one covariate.
For differences between pure and mixed pine stands (H6) at 75 cm, Márialigeti et al. (2009) give a possible explanation. They found that some terricolous species are correlated with the proportion of pine in the vicinity, as they are favoured by ‘better light availability, lower litter cover, and acidic soil conditions’ (Márialigeti et al. 2009). This could possibly explain the decrease in abundance when going from pure to mixed pine stands. To test this hypothesis it would be necessary to analyse our data for composition alongside the abundance data.
It is hard to analyse abundance values without having access to information on the composition. Further explanations for our results would need to refer to specialist species.
4.3 Critiques and further work
Further work is necessary in order to take advantage of the huge data base of information regarding numerous variables that is now available for further analysis. Further studies can take place in order to test the most interesting covariates as main explanatory variables.
19 a) Data for species composition
The first step for further work should be to analyse the composition of these populations, in order to complete the study of all aspects of biodiversity. It is difficult to analyse abundance values without comparing with composition. We were not able to find significant differences in both SR and abundance for a same group of bryophytes with further prevents much conclusive analysis to take place. Analysis of composition of bryophyte species, including a study into the frequence of both different species and different ecological groups per host tree, should be effectuated. This may help to ‘explain the mechanisms behind the tree species-biodiversity relation.’ Many bryophytes are host tree specific (Kiraly et al. 2010), which means that stand structure including a diverse composition could be important. It would be interesting to create models comparing the effect of host trees against physio-chemical factors for different specialist bryophytes, in order to ascertain whether the host tree species really is the most important factor at tree level.
c) Sampling design
Our data was perhaps limited by the small scale of the study. We only had seven plots for each stand type, and trees within each plot were pseudo-repetitions rather than independent values which is possibly the reason for under-dispersion of our results. Another problem was the fact that our sampling design didn't allow us to compare the overall diversity of mixed stands with that of pure stands at plot level. This was because the sampling effort differed between pure (three trees per plot) and mixed stands (six trees per plot). We were limited to comparisons of diversity associated with just one species (oak or pine) at plot level. However, it would have been possible to compare pure and mixed stands either by resampling our data (random draw of three trees from each of the mixed stands), or by accumulation curves (accumulation curves for SR over the seven plots for pure oak, pure pine and mixed plot types).
pH and conductivity of the above-ground water supply (throughfall and stemflow) was the most frequently occurring covariate in the improved model. We can hypothesis that the more acidic water supply associated with pine trees has a negative effect on the bryophyte SR and abundance. Other important covariates seem to be those associated with the stand density: BA and RDI. For epiphytes, our findings correlated with literature, and the larger the trunk diameter, the higher the SR. The correlation of higher BA with increased groundfloor abundance in pure pine stands compared to oak was contrary to other study results, and needs
20
to be investigated further. To test these hypotheses we would need to choose plots balanced for the variable.
d) Modelling
For unexpected results, such as the increase in bryophyte abundance in pure pine stands being correlated to an elevated BA, it would be interesting to create models with combinations of covariates. This would allow us to identify further interactions within the stand.
Barbier et al. (2009) discussed the usage of statistical models in the analysis of biodiversity-tree species relations, and stated that ‘the more the analyses target specific groups or species, the more we need non-linear statistical models. More research needs to be done on the type of models used for our analysis.
It would be interesting to study the understory vegetation associated with different tree species. Aude and Poulsen (2000) found that the shrub and sapling layer is important for epiphytic bryophytes, which can create a stable and humid microclimate by decreasing the effect of wind and desiccation, and Márialigeti (2009) found that understory layer is strongly correlated to the SR and abundance of ground floor bryophytes. A study by Nemati et al. (1994) on the understory cover in varied levels of mixed pine and oak ecosystems found that herbaceous cover response has a multivariate relationship with the overstory variables. In conclusion, the variation in understory cover between mixed and pure populations could have an effect on both ground floor and epiphytic bryophytes.
5. Conclusion
In the studied region, our results show interesting differences in bryophyte diversity at tree level between pure and mixed oak and pine stands. SR certainly seems to tend towards higher values in mixed populations for pine trees, and oak trees in mixed populations generally had a higher SR than pine trees. Bryophyte abundance at ground floor level also varied. Most interesting perhaps in terms of the aim of the study, and certainly with the largest number of significant results, was the difference in diversity on pine trees between pure and mixed populations.
In terms of prognosis for the forest management it is difficult for the moment to formulate a final conclusion, as we are lacking information for bryophyte composition. However, our e) Further field work
21
analysis does show that bryophyte SR and abundance seem to depend mainly on the host tree species, which suggests that the juxtaposition of pure stands does not differ from the mixture of species within the stands, from a plot level point of view.
Acknowledgements
First and foremost: Marion, your knowledge and encouragement was invaluable throughout the placement. Thank you for always having the time for me.
Yann, how would we have got through that field work without you? Always enthusiastic (‘it’s grésil, not grêle!’), always ready to offer advice.
Yoan, Patrick, Nathalie, Frédéric A. and Frédéric G., your generosity and availability never ceased to amaze me.
To the super troopers, whose kindness and good humour I will never forget. The Day when the Sun was Shining and we Ate Outside. The night when you all surprised me. Slack-lining and tree-climbing. Cake-baking and alcohol not-drinking. Etienne, my everything-friend.
Good old Carl, without whom I wouldn’t even have made it to work. You brightened up our lunch breaks. Always cheerful, always a story to tell (I’m still worried about those chicks), always there to lend a hand.
Yannick, my first confidant, and whose spirit rests on in Nogent.
Thank you to everyone: for your warm welcome, for your support, and for helping to make ‘the domaine’ my second home.
… And for the night when all was lost, and the gosling shepherd saved me from the depths of despair. It would have been impossible not to smile.
1 References:
Ahti T., 1977. Lichens in the boreal coniferous zone. Seaward MRD (ed) Lichen ecology.
Kjbhn,i45-181.
AFNOR, 1996. Qualite des sols. Association Francaise de Normalisation, Paris. 534.
Aude, E. and R. .S. Poulsen 2000. Influence of management on the species composition of epiphytic cryptogams in Danish Fagus forests. Applied Vegetation Science. 3:81-88.
Augusto, L. and J. Ranger 2001. Impact of tree species on soil solutions in acidic solutions. Annals of Forest Science. 58:47-58.
Barbier, S., R. Chevalier, P. Loussot, L. Bergès and F. Gosselin. 2009. Improving biodiversity indicators of sustainable forest management: tree genus abundance rather than tree genus richness and dominance for understory vegetation in French lowland oak hornbeam forests. Forest Ecology and Management. 258:S176-S186.
Barbier, S., P. Balandier, and F. Gosselin. 2009b. Influence of several tree traits on rainfall partitioning in temperate and boreal forests: a review. Ann. For. Sci. 66:602
Bardat, J. and M. Aubert 2007. Impact of forest management on the diversity of corticolous bryophyte assemblages in temperate forests. Biological Conservation. 139:47-66.
Barkman, J.J. 1958. Phytosociology and ecology of cryptogamic epiphytes, including a taxonomic survey and description of their vegetation units in Europe. Van Gorcum & Comp. N.V., Assen..
Ceschin, S., M. Aleffi, S. Bisceglie, V. Savo and V. Zuccarello. 2012. Aquatic bryophytes as
n iecological indicators of the water quality status in the Tiber river basin (Italy).
kjnkjiEcological Indicators 14 :74–81
Chang, S.C. and E. Matzner 2000. Soil nitrogen turnover in proximal and distal stem areas of European beech trees. Plant and Soil. 218:117-125.
Cleavitt, N.L., A.C. Dibble and D.A. Werier 2009. Influence of tree composition upon epiphytic macrolichens and bryophytes in old forests of Acadia National Park, Maine. Bryologist. 112:467-487.
Cronan, C. and W.R. 1983. Canopy processing of acidic precipitation by coniferous and hardwood forests in New England. Oecologia. 59:216.
Dzwonko Z, Gawronski S. 2002. Effect of litter removal on species richness and acidification
jhchkiiiof a mixed oak-pine woodland. Biol Conserv 106:389–398.
F. Beniamino, J.P.a.P.A. 1989. Soil acidification under the crown of oak treesI. Spatial distribution.
Farmer, A., J. Bates and J. Bell 1991. Seasonal variations in acidic pollutant inputs and their effects on the chemistry of stemflow, bark and epiphyte tissues in three oak woodlands in N.W. Britain. New Phytologist. 118:451.
2
Felton, A., Lindbladh, M., Brunet, J. and Fritz, Ö. 2010. Replacing coniferous monocultures with mixed-species production stands: An assessment of the potential benefits for forest biodiversity in northern Europe. Forest Ecology and Management. 260:939-947. Fiedler, H.J., W. Nebe, W. Hoffman, P. Kruger, K. Dreger and H. Rollig 1981. Relationships between site and spruce growth in a region near the Baltic Sea. Archiv fur Acker und Pflanzenbau und Bodenkunde. 25:245-255.
Frédéric André, M.J., Quentin Ponette 2008. Effects of biological and meteorological factors on stemflowchemistry within a temperate mixed oak–beech stand. Science of the total environment.72-83.
Fritz, Ã., M. Niklasson and M. Churski 2009. Tree age is a key factor for the conservation of epiphytic lichens and bryophytes in beech forests. Applied Vegetation Science. 12:93-106.
Glime JM. 2007. Physiological Ecology. Bryophyte Ecology. 1.
Gustafsson L, Eriksson I (1995) Factors of importance for the epiphytic vegetation of aspen
jhglgliui(Populus tremula) with special emphasis on bark chemistry and soil chemistry. J
iiiiiiiiiiiiiAppl Ecol 32:412–424.
Hauck M., T. Spribille. 2005. The significance of precipitation and substrate chemistry for
kjhkjhjepiphytic lichen diversity in spruce-fir forests of the Salish Mountains, northwestern
dsfkjh Montana. Flora 200 (2005): 547–562
Humphrey, J. .W., S. Davey, A. .J. Peace, R. Ferris and K. Harding 2002a. Lichens and bryophyte communities of planted and semi-natural forests in Britain: The influence of site type, stand structure and deadwood. Biological Conservation. 107:165-180.
Kiraly, I. and P. Odor 2010. The effect of stand structure and tree species composition on epiphytic bryophytes in mixed deciduous-coniferous forests of Western Hungary. Biological Conservation. 143:2063-2069.
Koptsik, S., N. Berezina and S. Livantsova 2001. Effects of natural soil acidification on biodiversity in boreal forest ecosystems. Water Air and Soil Pollution. 130:1025-1030. Kuusinen, M. (1996) Cyanobacterial macrolichens on Populus tremula as indicators of
dskjjjjjjjforest continuity in Finland. Biological Conservation, 75: 43-49.
Márialigeti, S., B. Németh, F. Tinya and P. Ódor 2009. The effects of stand structure on ground-floor bryophyte assemblages in temperate mixed forests. Biodiversity and Conservation:1-19.
McGee, G. and R. Kimmerer 2002. Forest age and management effects on epiphytic bryophyte communities in Adirondack northern hardwood forests, New York, U.S.A. Canadian Journal of Forest Research. 32:1562-1576.
Nemati, N., and H. Goetz. 1995. Relationships of overstory to understory cover variables in a
3
Peck, J.E. 1997. The association of commercially harvestable bryophytes and their host species in northwestern Oregon. Bryologist. 100:383-393.
Puckett, L.J. 1991. Spatial variability and collector requirements for sampling throughfall volume and chemistry under a mixed-hardwood canopy. Canadian Journal of Forest Research. 21:1581-1588.
Pryor S., Barthelmie R., 2005. Liquid and Chemical Fluxes in Precipitation, Throughfall and Stemflow: Observations from a Deciduous Forest and a Red Pine Plantation in the Midwestern U.S.A., Water, Air, & Soil Pollution, 173 (1-4): 203-227.
Raabe, S., J. Muller, M. Manthey, O. Durhammer, U. Teuber, A. Gottlein, B. Forster, R. Brandl and C. Bassler 2010a. Drivers of bryophyte diversity allow implications for forest management with a focus on climate change. Forest Ecology and Management. 260:1956-1964.
Raabe, S., J. Müller, M. Manthey, O. Dürhammer, U. Teuber, A. Göttlein, B. Förster, R. Brandl and C. Bässler 2010b. Drivers of bryophyte diversity allow implications for forest management with a focus on climate change. Forest Ecology and Management. 260:1956-1964.
Schmitt, C.K. and N.G. Slack 1990. Host specificity of epiphytic lichens and bryophytes: a comparison of the Adirondack Mountains (New York) and the Southern Blue Ridge Mountains (North Carolina). The Bryologist. 93:257-274.
Slack, N.G. 1976. Host specificity of bryophytic epiphytes in eastern North America. Journal of Hattori Botanical Laboratory . 41:107-132.
Vallet, P., and T.Perot 2011. Silver fir stand productivity is enhanced when mixed with Norway spruce: evidence based on large-scale inventory data and a generic modelling approach. Journal of Vegetation Science:1-11.
Vellak, K., J. Paal and J. Liira 2003. Diversity and distribution pattern of bryophytes and vascular plants in a boreal spruce forest. Silva Fennica. 37:3-13.
Virtanen, A.E.J., M. J. Crawley and G. R. Edwards 2000. Bryophyte biomass and species richness on the Park Grass Experiment,Rothamsted, UK. Plant Ecology.
Wallrup, E., P. Saetre and H. Rydin 2006. Deciduous trees affect small-scale floristic diversity and tree regeneration in conifer forests. Scandinavian Journal of Forest Research. 21:399-404.
Weibull, H. and H. Rydin 2005. Bryophyte species richness on boulders: relationship to area, habitat diversity and canopy tree species. Biological Conservation. 122:71-79.
